一种软判决下的本原BCH码盲识别方法

A Blind Recognition Method for Primitive BCH Codes in Soft Decision Situations

为解决现有BCH码识别方法容错性较差的问题,提出了一种软判决下的本原BCH码盲识别(SDBR)方法。首先,将截获数据进行解调软判决得到比特序列和对应可靠性信息,然后对比特序列进行码字划分,再由软判决可靠性信息建立码根可靠性系数,以此计算不同码根出现的概率,并引入Kullback-Leibler散度来确定码长;其次,定义公共码根可靠性统计量并建立二元假设检验,在不同本原多项式下对公共码根进行判定;最后,利用公共码根连续分布特点识别本原多项式,进而由所有公共码根计算生成多项式。仿真结果表明:SDBR方法在信噪比大于7dB时能有效对常用本原BCH码进行识别;与基于码根信息差熵的方法相比,容错性提升了1.8dB。

A soft decision based blind recognition（SDBR）method for primitive BCH codes is proposed to solve the problem that existing recognition methods have low error-resilient capabilities.Firstly,bit sequences as well as the corresponding reliability information are obtained by soft demodulation of the intercepted data,then code words are divided and a reliability coefficient of code roots is established with the reliability information to calculate the occurrence probability of code roots,and code length is estimated by using Kullback-Leibler divergence.Secondly,reliability statistics of common code roots are defined and a binary hypothesis test is built,then common code roots are verified under different primitive polynomials.Finally,the right primitive polynomial is recognized by using the continuous distribution characteristics of common code roots,and a generator polynomial is calculated from all these code roots.Simulation results show that the SDBR method effectively recognizes the commonly used primitive BCH codes when the signal-to-noise ratio is above 7dB.A comparison with the roots information dispersion entropy based method shows that the SDBR method improves the error-resilient performance by 1.8dB.